LINUX Unplugged - Episode 189: Das Boot Manager | LUP 189

Episode Date: March 22, 2017

Bulletproof Linux Kernel upgrades might be near, Kodi gets a real Netflix Plugin & the dirty, stinky, no good, obvious, elephant in the room around desktop Linux.Plus why Bcachefs might be Linux’s next hit filesystem, Mozilla's Obsidian & more!

Wes, there is a conspiracy afoot. It appears that Office 365 OneDrive looks at user agents to determine performance. This is breaking news on the r slash Linux subreddit, which everyone knows is the bastion. This is CNN breaking news. CNN-level style reporting here, but this is something to consider from Tornator. He says, a few weeks I've been running Linux Mint 18.1. I got it on my lappy. I'm still in school. I started a new project. We have a project where we use one drive on an Office 365.

When he uses Firefox 52 and goes online, he gets it in compatibility mode. Sadly enough, he says, I experienced a lot of performance issues with that. He says, I have a desktop running Windows 10 and the same Firefox 52 web browser. I tried working on the document, and I see no, I see if the problem persists, but it does not. There was no problem. I began thinking about trying a few different things, so he began switching the user agent under his Mint install to make it look

like it was running on Windows NT 6.1, 64-bit version of Windows, but still Firefox Gecko. And after changing the user agent, he says the performance problems were resolved. The UI of OneDrive worked flawlessly. The only thing that was changed in the user agent was the OS. He left all the version stuff of Firefox the same.

He thought maybe it was a random occurrence, so then he changed it back to the regular agent. Problem comes back. Have we discovered a massive controversy here, Wes? Or do you think it could just be a testing thing? Like, we didn't test that one very well, so we'll put it in compatibility mode. I am inclined to think that you should more often attribute things to bugs rather than malice. I would imagine so in this case. Yeah, especially in a case like this where I wouldn't expect Microsoft to put a ton of resources into researching and testing the Linux side of things.

So this should work, but we haven't tested it. So slap compatibility on there. That does look pretty plausible. All right. Darn it. Conspiracy busted right here on the Unplugged program. This is Linux Unplugged, episode 189 for March 21st, 2017. Welcome to Linux Unplugged, your weekly Linux talk show that's alive and strong.

My name is Chris. My name is Wes. Hello, Wes. Hello, sir. You know what,. My name is Wes. Hello, Wes. Hello, sir. You know what, Wes? Today is a great day. We have so many fun topics to cover. Oh, yeah.

For me personally, there's some great ones in here, too. But I know for some of the other projects the show's just talked about in the past, we got some great updates. And then, and then, I am really excited. We have a great attendance in the virtual lug today. And I know one of the main topics we're going to get into later in the show. I don't want to say too much because I don't want them to spoil it. But we have a really

great conversational topic about long-term future desktop stuff that's sort of an elephant in the room kind of stuff that we're going to be talking about. We'll also get an update from Ike on what the Solus project is up to. Some new code has been pushed and I got a few questions. But obviously, obviously we're going to start with a bunch of updates that are totally, totally rocking.

So, Mr. Wes, we've been supplied with – are these an ale, technically, or are these considered a cider? It's a wheat ale. Oh, it's a wheat ale. Okay, so to help us, to help grease the wheels of Linux Unplugged, Wes has supplied us with a passion fruit kicker. Oh, yeah, a wheat ale with passion fruit built in. So that is, that'll be what we're drinking along. Here you go, cheers.

Cheers, Wes. Cheers to 189. Yeah, and to Linux. And to our mumble room. Oh, speaking of that mumble room, let's bring them in. Time-appropriate greetings, virtual lug. Greetings, progress. Hello. Hello. that mumble room let's bring them in time appropriate greetings virtual lug greetings hello hello i am very excited a little bit of uh late breaking news as we go into the show today um those of us that run cody are going to be very very excited by this news

it looks like a proper netflix plugin could be coming coming to Kodi soon. In fact, if you're on Kodi Android, you can get your hands on it already. So this is not a sanctioned thing. I'm going to just get that big caveat out of the way. But after years and years of using different workarounds, there is now sort of this legit Kodi native Netflix plug-in add-on being developed. It's open source, so that's kind of a big deal right there.

Just this plugin itself being open source is nice because not only does it mean maybe we could get off Android Kodi, but it could also do it from architecture. Yeah, exactly. So it's interesting because it's based on Kodi's input stream binary add-on extension, which can serve as an input stream for Kodi's video player, which is new since Kodi 17, which has been a massive update. The add-on enables playback of the DRM-protected content

without having to break the DRM distribution chain that Netflix uses to encrypt its video stream. Now, I think that's important because it means that Netflix might be content pretending like this doesn't exist because you're not doing anything improperly. You're not busting the DRM. It only works with Kodi built for Android, but it should be possible to port it to Linux. Right, like as long as it's not enabling workflows that break your user agreement or make it really easy to copy the content. Make it easy to pirate. Yes, exactly.

Then maybe they will just. Uh-huh. So this slide. It is up to pirate. Yes, exactly. Then maybe they will just – Uh-huh. So this – But this slide. It is up on – but this is a messed up post. The Kodi 18 isn't – anyways, it says the prereq is Kodi 18, but the Pharonix post says the prereq is Kodi 17. I would guess it's 17 since 18 isn't a downloadable thing right now. But for me, using the NVIDIA Shield Android TV and having Kodi on there,

and right now I use Kodi to also play Plex media. Yes, right. I find that every now and then when I back up something, there's like a flaw in the file that I don't catch until I've tried to watch it. And the standard Plex player just dies. It hits a couple of dropped frames or a dropped Plex player just dies. It hits a couple of drop frames or drop audio pack and it just dies. Kodi will just and then keep going.

Go. And it's faster too and plus you get all the other benefits of Kodi. You don't have to leave Kodi. So now I can add one more thing to that. That's where it would change things for me is I could go from before I have to rely on things like a Chromecast or maybe it's built into a Smart TV or I have a third-party device. If it could all just be right in my Kodi appliance, it really would make it more of an appliance

to me because I do, I mean, or the people that use it in my life and a lot of us use Netflix. Yeah, otherwise you're always jumping out of Kodi to go watch Netflix eventually. Now, we have Mr. Kernel Linux joining us this week and he and I did like this accidental versus like cheap android tv box versus the nvidia shield which is kind of on the higher end i've since bought two more nvidia shields since that review i really like the product but i wanted to check in with you are you still using uh are you still using that android tv review to ages ago or have you all

have you rolled all back to the western digital lives nope so i have still in the house the majority of the um of the media players are still the western digital tv lives the box you're talking about is the matricom uh m-a-t-r-i-c it's like 99 bucks right yeah it's a little cheaper than that and i have the i have the q2 which comes with the the built-in ir uh connector as well as uh wired ethernet um and so far i've had not i have not had any problems. I'm still using the Q2 downstairs in my shop. So basically every day when I'm working on something, I just have it kind of up playing some media. And I really, really like it. I only have two hits. I found

once I tried to take it with me to the lake, and the Western Digitals are all 12-volt DC, so you, you know, living in an RV, you can understand why that's amazing. And the Q2 is 5 volts, so I have to use either an AC inverter or I'd have to use, like, a voltage step-down transformer. And then the other thing I don't like about it is there's no – the Western Digitals are very square, so they mount very nicely to the wall. And the Q2 is kind of a – I'm surprised how many boxes do that. Yeah, that's weird, right? That is a thing. They do make a bracket for the Roku,

and I was planning on using that. But the fact that you have Android means that it works with Netflix, means it works with Kodi, and the ability to be that flexible is actually worth overcoming some of the other shortcomings. That's exactly been my takeaway. Ike, I saw you in there saying,

Ha! Can you explain a little bit of that? I mean, I saw a lot of that. It seemed like you were going, Ha! Using a player. I just use my web browser like a man. Explain yourself.

Well, yeah, I mean, that's pretty much it. I like that I do this massive characterization. Yeah, that's about right. I have YouTube and I have Netflix in a browser. I mean, what else do I need? I don't have downloads. I think for me, and I think one of the reasons I talk about Cody more is when I moved into Lady Jubes, I made this huge shift into consuming a lot of my video content via the television instead of on my desktop.

And it's a totally different experience for me. Like it is a legit way for my, my fiance and I to sit down. Right. Especially when you're sharing it with people. Yeah. And the family too. Like it is, it has become like our primary way of, we don't have television. We barely get any, where I'm at right now, I get decent reception, but usually I don't

hardly have any reception. So for me, it's, I, I, yeah, obviously if I'm at my desk, I'm using Chrome. Yeah. I'm not using Kodi. I'm using Chrome. But if I'm on the, if I'm on my TV, I want something that is appliance level, reliable and fast. And Kodi is the only one that delivers that. And it has, you know, like it integrates well, if you want to tie in like an IR remote or other things that will be feel first class in the TV ecosystem. Yeah. Yeah. Yeah. I mean right now I have it all tied in with the Echo. Oh, right.

Yeah. There you go. Yeah. Something. Don't tell Noah. But I just walk in. I tell it to turn on the TV and it just does all that stuff. All right.

Well, let's move on because I know not everybody cares about Kodi. And this next one, I just wanted to take a moment. There's two pieces of kind of – I don't know if malware is the right term. It's not because the first one is ransomware and the other one is an exploit. But I just thought both these stories were interesting. And so I wanted to throw them together really quick. First, I love this. There's new ransomware out written in Python and it's called Kirk Ransomware.

Yeah, it's Kirk Ransomware. And it's the first ransomware really that uses the – are you familiar with this Monero payments? It's like an alternative. It's an alternative to Bitcoin. What kind of Bitcoin? What? Oh, yeah, okay. It's an open source cryptocurrency created in April 2014. Yeah, I don't know why I've never heard of it. No, I haven't heard of it either. Yeah, but I'm sure...

You know what? I'm going to read on, but I have a little bit of bacon about this. I kind of suspect something here. Oh, you got the bacon? Yeah. So it's written in Python, so it seems like it could probably be moved over to Linux and, you know, screw up your home directory if you're not careful. But there was one little bit of tidbit that I liked about it. It is not known how it's being distributed, so don't worry. This is not something that's like a big deal.

Some of the details are interesting. It masquerades as a network stress testing tool called low orbital ion cannon, which we've heard of before. So what happens is people go out and they look for like, I want to do a network stress test or I want to just be a dick to somebody and they go get low orbital ion cannon thinking that, well, this is the Python script that does it. Turns out, no, actually it encrypts your home directory.

The Kirk ransomware will then generate an AES password that will require the user to put in the unlock key to decrypt the files. But this is the best part. This is the best part of all of it. The decrypter tool? Do you have a guess what it's called? It's called Kirk ransomware. What do you suppose the decrypter tool is called? I'm really hoping for something with con or Spock. One of those two. It's Spock.ansomware. What do you suppose the decryptor tool is called? I'm really hoping

for something with Khan or Spock. One of those two. It's Spock. The decryptor is called Spock. Yeah, yeah. You run the Spock decryptor and they supply that

to the victim once the payment's made. Nice. And I mean, the live long and prosper thing after you decrypt. Yeah, you like that? That's nice.

That's nice. They got a screenshot of that. Really? Some production quality? Like, you know, they wanted the appearance to be nicer as they're making you pay them. It's funny. It's mean.

It's, I think, a promotion for the cryptocurrency. Yeah. I think somebody created this. What? Like, why not con? Yeah. Well, I guess because Spock is saving the day, right? Because con would be destroying your files and Spock is rescuing your files.

Yeah, so if it didn't, if it, like, fake decrypted and just deleted them, then it would be con. I really feel like this is a promotion for a cryptocurrency. Like, hey, how do we get some movement in the market? You know what we do? We get people to have to go buy some. Yeah. Yeah. Not only does he get the name out there, but then he gets-

I've already bought six Moneros. That's not true. I'm mining Moneros right now, dude. Now, Kirk will ruin your plans. So maybe we should like create a fix for this ransomware that will – maybe that will be called Khan, the fix for the ransomware. Right, right. I think – or the ransomware should have been called the Wrath of Khan. All right.

And then just one other thing just to follow up from a story we covered recently, that CIA Vault 7. Well, it turns out now that with some of the – it's funny because WikiLeaks has been getting a hard time for releasing code that makes us more vulnerable. But I wish they would have released more code so we could have built better tools to detect where CIA tools are being used. But there have been some people, some antivirus companies that have taken some of the WikiLeaks releases and generated pattern recognition to detect the tools and stuff like that. So there is some stuff that's coming out. And there's this one. A simple command allows the CIA to commandeer 318 different models

of Cisco switches. The bug relies on a Telnet protocol relies on the Telnet protocol and yeah, over 318 switches. Alright, what's the command? Yeah, right? I don't know. Because all they say is the bug resides in

the Cisco cluster management protocol, which uses the Telnet protocol to deliver signals and commands on the internal network. So I don't, I don't, ours is not really giving you the command here in the article. Interesting though. I thought, you know, just wanted to follow up on the Vault 7 stuff and

also include it maybe with the Star Trek malware, or I'm sorry, ransomware. Oh, that's beautiful. Or horrible. Yeah. Now, William, I see you're back in the room and your timing couldn't be more perfect because there is one big update on a project that you've gotten me really excited about. And that is Bcash FS. I think a new version has shipped and lots of stuff's been going on.

I think the biggest milestone has been that all of the breaking on-disk format changes are landing. Like, all that stuff's getting out of the way. That should be over now, in theory. But there's also some other big features that have rolled out. Where should we start when we're talking about the new Bcache? FS. I think the most interesting thing is the on-disk encryption stuff that we're doing. Yes.

With Bcache FS. That's great, because you build that right into the file system and so now you'll have competition in that space and I think what's even better is that not only do you get encryption but you also get verification so that someone can't just come along and write bytes to the underlying file system and corrupt your data without you knowing about it. Because with

something like DMcrypt, you just get encryption it doesn't actually check to see if what you wrote is authentic. Okay, I hadn't even thought of that. And then also, can you tell me about the wonders of the new backup superblocks? I mean, that just brings it up in the line with what ButterFS and ZFS do, right? So you get a bunch of superblocks distributed across the file system. That way, in case one of them gets corrupted, you still have backups to go back to. Because before, I think on the version that I'm using on my laptop, you just have one super block.

So the moment that one thing gets corrupted, you lose everything. So is this new in-node format, is this one of the larger sort of practical changes in the new release? It's not necessarily interesting to end users, but it is a little bit more compact, and it does allow him to add more features to the file system over time. So I think it's actually a good thing. Oh, and also I like this line,

being able to fit all of your metadata in RAM. Smiley face. Yes. That is very interesting. A lot of them are just small underlying changes that make it more performant. Userspace FSCK, that seems kind of nice. Yeah, that is nice.

Yeah, actually having the FSCK already done before it's even integrated into Linux is probably a huge selling point because neither ZFS or ButterFS had that. How do you feel about the file system migration stuff? So apparently it works flawlessly. I've never tested it. I actually am not using the new version yet.

I'm still on the old disk format. I mean like somebody wants to bail on ButterFS because it's garbage or somebody wants to upgrade from Extended 4 or somebody wants to purge ZFS from their life. It's fantastic because it'll take any file system, it'll build an image, and then you can rewrite that image over once you verify the image is correct.

Obviously, this means that you're going to store the data twice, so you have to be using less than 50% of your file system currently. But it is nice if you have a fairly small amount of data and you want to migrate your root FS over or something. And you could also... Oh, yeah, okay. So, could you

actually boot from a Bcache FS file system? Yeah, yeah, I. So could you actually boot from a BcacheFS file system? Yeah, yeah, I'm booting my laptop. I'm running the whole thing on BcacheFS. Oh, I thought that wasn't a thing. The EFI partition has to be read by the EFI. So it's FAT32. But otherwise, it's all Bcache.

It's built right into the kernel. Unfortunately, that doesn't mean you have to build his Linux tree. Right. I maintain my own with the most stable patch set. He actually just keeps it at the base 4.9 and just keeps adding patches on top. But I merge in the stable versions every now and then as soon as they come out. So I've just been keeping with that. It's a fairly up-to-date kernel.

It's still 4.9. It's not 4.10 or the RC. But he pushes it forward every now and then. So Kent's assertion here is that RAID 1 and RAID 10 are... RAID 1 and RAID 10 potentially work. It's not very well tested, is my understanding, and I'm not 100% sure the recovery code is even in place. So it'll distribute the blocks across multiple disks,

but beyond that, I'm not really sure how well you'd be able to recover your data if stuff went bad. I know it is doing checksumming, so it is aware of disk errors. I'm just not sure what the replication code does in the case of seeing an error on one disk if it copies the crunk over from the other disk. What is the tooling like around the replication support? I'm familiar

with ZFS send, obviously. Is it in line with it? That's sending and receiving file system images more or less. This is different than that? This is the more or less. This is different. Okay, this is different than that? This is the RAID levels. This would be like running mirrored mode in CFS or running RAID 1 in BinaryFS.

Okay, okay. I think eventually he's going to grow send receive support. I just don't think that's a major feature, major milestone on his plate right now. Right. Man, this is sounding like an awesome frickin' This is sounding like my future Linux file system right here. It'll be really interesting to watch as it

gets closer to being ready to be upstreamed. I think someone said deduplication in the chat room. No, there's no deduplication yet. I assume that will also be something that's pretty trivially possible at some point. I think the hardest thing is going to be getting snapshot support. Oh, yeah. Why is that? Oh,

just given the structure of the file system using B-trees, it's fairly hard to do snapshots correctly. Oh, okay. I think in ButterFS it's actually fairly complicated logic. Oh, because that was my next question, is it seems like that was something that ButterFS sort of rolled out with.

Isn't this not a copy on Sorry, is it not a copy-on-write file system? It is a copy-on-write file system. Then, normally... Copy at that point? Yeah, you copy it

and then just stop... Sorry. You just stop writing where it was writing and write somewhere else. Sort of. I know that's a gross oversimplification. That's a gross oversimplification and I don't think those are the underlying issues with doing snapshotting well.

Hmm. You know what? I would love to know more about that. Maybe we can talk more about it sometime. I'm not necessarily very familiar with the topic but I think the hardest part is actually purging old snapshot data. And keeping ref counts correctly and those sorts of things. And then rebalancing the B-tree.

This is a project we've been watching now for, I don't know, it feels like it's been a little while now. It's been going, yeah. He announced it, I think, in August or June in that time frame, somewhere in the summer. You've been running it for how long? I've been running it since he announced it, basically. And I haven't reformatted once. Love it, dude. And I will say there's probably a bit of corrupted stuff on there.

We'll just hit a little. Wow. But it seems to be running okay still. I think one of the nicest things is the long tail latency is fairly low. So all your operations perform incredibly quickly when you're just trying to do small accesses. Oh, man, that's nice. Yeah, random performance is very good, assuming that the underlying device is also decent.

Right, like a modern SSD or something. Yeah, yeah. Which is what you want. Yeah, because the problem with ButterFS, I've noticed, is you get a lot of hangs doing certain operations every now and then. Like, it'll just go to sleep for a second and come back. You're like, yeah. ZFS is better with that.

Especially under different conditions. There's some conditions where that seems to be worse too with ButterFS. It's not a well understood thing. It's a very complex code base. So yeah, there are tons of edge cases where you can run into really long tails and Bcache does as much as it can to not run into those

long tails for reads and writes. So it tries to avoid doing most of the work in the critical path. So if I get this kernel and then maybe if I compile that Wire. In the synchronous path. Maybe if I compile that WireGuard module too, I can just be the ultimate kernel hipster. So you're just going to switch to Gen 2? Got my WireGuard.

Nice. Living in the future. You do? It's like William's like, yeah, that's my setup. What disk are you using, William? I'm using NixOS. Oh, yeah, that's right. That's why you're one of my favorites.

We've got to do an episode on that. Yes, we do. We can do an episode. I can show you NixOS in production. Awesome. Yeah, write that down. I think I want to do that. I'd love to know about it. We'll figure it out. Yeah, cool. Alright, well, let's talk about Solus here in just a

moment, but first, I would like to say a big thank you to Linux Academy. I am so happy with the growth Linux Academy has seen and the pickup they're getting, the partnerships they're establishing now. It's one of those things where if you signed up like a year ago, like every single week, something new and just constantly improving, it is so, so valuable. But if you're new to it, I would encourage you to get started at linuxacademy.com slash unplugged. You get a seven-day free trial.

That is going to get you acquainted with, like, the essentials of this platform. It's a platform to learn about Linux. It's a platform to advance your career and go into a new field. It's a platform to double down on the area that you work in. on the area that you work in. It is so well built because it was designed by Linux enthusiasts, Linux educators and developers who came together and said,

all right, how can we advance Linux? And I had this conversation with their CEO before they were even a sponsor when we just started flirting about... Feeling it out, feeling it out. No, that's a whole different... Hold on, hold on to a different picture. We were just kind of talking.

He's talking at first. And he said, you know, we watch JB. But I don't understand how you make any money at it. And I said, well, we don't make very much. He's like, yeah, well, we think we might have a way to advocate Linux, make money, and actually improve the workforce in a way that is like a tool that was never available when you were in the workforce. I'm like, oh, yeah, sure, Anthony. Yeah, that'll happen. And you know what?

A couple of years later now, it is unbelievable what they've accomplished. Self-paced, in-depth video courses on every Linux, cloud, and DevOps topic. And they're not afraid to branch out into areas that you need. It's very impressive. Hands-on, scenario-based labs that give you experience on real servers. Instructor mentoring. That means an actual human being that knows what they're talking about. An instructor that will work with you when you need help.

That's phenomenal for the type of courseware they're covering. This is unique to the Linux Academy. They have lab servers that spin up on demand as you need them, a course scheduler for your busy schedule. I would like you to go check them out at linuxacademy.com slash unplugged. That supports the show, and from there you can sign up for a free seven-day trial. The other thing that's nice is even when life gets busy, like say it's the holidays, right?

Life is super busy. They have nuggets that can be just a few minutes long where you can just go in and get a little value that week. They have availability planners that can build courseware for your busy schedule. And on top of all of that, you can just log in and just get snapshots of what your progress is at so you can kind of just keep track of how you're doing. You get time estimations for every major topic. You want to learn Python. You want to learn Ruby. You want to learn Ruby. You want to learn PHP.

It's going to take you six hours, four hours, seven hours. It's like quantifiable. Yep, it's great. Like I know – I don't know how to wrap my head around learning Ruby. I know how to wrap my head around six hours. You have to go find all the documents, then do it. Oh, for sure. This is all just right there.

Right. You have an idea of how long it's going to take. And for me, taking it down from I'm going to learn this nebulous topic that is massively broad and there's so many different opinions, taking it from something like that to I'm going to spend six hours and learn a thing is so much more – it's like I have so much more motivation to do that because it's a clearly definable object. To that same point, you can just browse through, right? You're like, maybe I want to learn Ruby.

By looking, you can see all the stuff you'll do after going through their resources. You'll have an idea like, okay, yes, I'll be able to accomplish my goal after six hours. Absolutely. Linuxacademy.com slash unplugged. Go there to keep us going. Support the show. Let them know you heard about it here and then sign up for a free seven day trial. And by the way, the community rocks because it's stacked full of Jupyter Broadcasting members. They fork like the study cards and stuff. So that way they get even better over time. It's nice. And they've got downloadable guides you can take with you and lesson audio, personal notebooks, tools that help you study iOS and Android apps as well. LinuxAcademy.com slash unplugged.

Thanks, Linux Academy, for sponsoring the Unplugged program. So Mr. Eike joins us again, which is – it's always a nice pleasure. Thanks for being here, Ike. I appreciate it because I know it's not like the middle of – it's not like the afternoon over there like it is here. So I always really do appreciate it. It's about 9 o'clock. Well, 20 to 10 now. I'm grokking from what's happened that you probably don't sleep much anyways because it looks like a lot has been going on with Solus.

It got Mate Mate 1.18. It got a new Linux kernel coming in. It got a new budget coming in. Tell me what's going on over there. Yeah, basically all that stuff. And lots of new apps, too, or updates. I mean, Mate 1.18 went in, I think it was the day after or maybe two days after.

I'm not even sure now. We got that into Solus. We got that tested very quickly and out to people. No breakages. I was very, very happy with that. We got the lib input support as well, so that's very nice. That's great. Oh, yeah.

So now people will have less touchpad problems. Really? That's super nice. Yeah, so there's been bugs about that on the tracker, longstanding issues, but Upstream solved them, so we have those in. One of our changes from Solus that actually landed for this release cycle was the action icon support within mate notification daemon oh nice it's a trivial little change but it just allows you to have icons instead of just plain text on there yeah and some applications like rivenbox express more actions when icons are available so it's a it's a more natural use of things like media players.

So we was happy to get something in, at least, into that release. We've got 4.9.16 kernel, I believe, we're on now. We're following the LTS tree for now until we got Clear Boot Manager integrated and then we can have switchable kernels. How is that going? That's actually going really well.

So the last bit of stuff that I was doing there was basically implementing proper namespacing on the EFI system partition. So the decision was made to namespace under the EFI directory. So for Solus it would be slash EFI slash com dot Solus project. And then all of our files will live

in there and we're going to want to add things like uh firmware update support in there using the fw upd project and i've pretty much finished the grub2 support for legacy booting so i need to test that on the laptop but once that's, it's pretty much close to a 2.0 and then that will be deployed first as a suicide test in Solus. Remind me, yeah, I mean, it seems like you're going to be, aside from Clear Linux, won't you be the only other distro that's shipping that as the bootloader? Yes, well, it's not the bootloader itself. The idea is you manage the bootloaders. Explain it to me because I barely understand.

So explain it to me. Yeah, so basically you want to have a safe and bulletproof approach to booting and the management of kernels themselves. Right, right. We've broken it down a few weeks ago, and I thought it was actually a pretty brilliant idea. It's an Intel project, right? Yeah, so basically bootloader tools

you can kind of trust them not always uh when it comes to things like case sensitivity it matters to linux it doesn't matter to fat32 so simple things like that if you was to go and rename some of your directories to some cranky mixed case names and then if you started to use things like a gummy boot or Boot Control, you're going to run into problems where it can no longer install or update because it's trying to create new named directories that already exist. So it started out at those problems and having a way of updating to a new kernel and retaining the last kernel without having to have lots of meta packages

depending on new versions of those packages to push the updates because you don't want to end up where you've got 47 different kernels on your system and no way to garbage collect them so i had to implement a policy whereby it knows if a kernel is booted which is very simple that's just a system d job to run the tool right it then needs to know what the new default kernel is it needs to know about different types of kernels and when to kill an old kernel so basically everything lives in user lib kernel and once once you call the tool everything happens in a single atomic atomic operation if you can't continue it's going to bail and it will try its absolute best not to murder your

system this is really an attempt to get as close as possible to bulletproof kernel updates, right? Yeah, that's the thing. So sometimes, I mean, we've all been there. You've updated kernel. You boot to the new kernel. Ah, crap. You know, it's completely trashed. Now, if you're using Solus or even Arch Linux, you know yourself the way the kernel updates work.

Your last kernel is gone. It's just completely gone now. You're on the new kernel you've had the carpet ripped out from underneath your feet and all your kernel modules are gone you've got a reboot so it's all those things now distributions like ubuntu they use meta packages to depend on the new package they put out right but then you still have all those old packages you still have the complexity of it you still have poor maintenance of the esp itself where you have something like slash vm liners how do you dual boot with that it's just not really good i mean we've dual booted for years with legacy boot purely because we all maintain

our own separate boot trees and we get away with that right when you start having a boot partition you start to get conflicts. When you've got an ESP, you're almost certainly screwed. So, you know, you need to be able to dual boot. You need to be able to have bulletproof updates and always, always, always do the right thing. So if something does go wrong, you can just boot back to the old kernel.

That's going to be a great feature. It's a great way to make sure, you know, feeling comfortable with updates, especially kernel updates. Yeah, I mean, imagine, just look, you know, feeling comfortable with updates, especially kernel updates. Yeah. I mean, imagine, just look at the hoops that Mint goes through to avoid kernel updates. I would much rather just know that I can boot right back into my last configuration if I need to. Yeah.

I mean, that's kind of the main thing. Now, the problem you have, like, if you use systemd boot, you know it's basically the same as gummy boot, right? Yeah. And Solus, a long time time ago forked gummy boot when we had the whole fiasco of systemd oh really swallowing uh gummy boot and my reasons there was the whole distro side of it not the actual boot side of it it was things like the case sensitivity because it's actually a case of case ignorance because linux cares fat does not care but if you try and create two files, uppercase and lowercase,

Linux is going to scream. But FAT, again, doesn't care. It would just fail to create the file because it already exists. Linux doesn't know it exists. So because of problems like this, and basically upstream systemd didn't care at the time, or being able to upgrade from GummyBoot 49, I think it was, to the first systemd boot, again, they didn't want to implement that logic,

and I wasn't particularly happy. So that's why my fork came about. But the stuff that went into that, that approach to the ESP and the maintenance of it, that's what Clear Boot Manager now does anyway. So I can now drop a fork, I can adopt SystemD boot, and I can have Cbm manage all these things the nicer bits that it does is it automatically generates and maintains the

loader entries on the esp so you probably seem like the loader entries star.conf files that you have what cbm does is it will emerge the upstream command line files which is on a per kernel basis so we could have like linux and linux lts command line in those yeah you can have etc command line files which is on a per kernel basis so we could have like linux and linux lts command line in those yeah you can have etc command line files as well which are then merged into it it will dynamically find the root partition and put the root equals part into there as well so you don't have to go looking for block id stuff anymore yes i hate that you know the pain with that and one of the nicest things it actually does um at the moment it's limited to draco but it can actually find out if you're running an encrypted root fs and put in the lux uuid in there for you which i've had problems with that with

grob and if you're manually creating entries with something like systemdbo it's very very hard to get all those things consistent so it does all of this stuff for you create entries just so you're safe all the time pop up bless you see kernel linux when i say i feel like solace is a is a different distribution even if it might be a smaller distribution i you know i think it's something you can't know as a regular end user maybe you can tell by certain external properties of a project but there's no way to really know, unless you listen to these shows, I suppose, like the heads behind these projects. Like, Ike, that's – you're obviously – you're solving a problem there with Clearroot

Manager that Red Hat is trying to solve, that Canonical is trying to solve. And this seems to be a very, very, very obvious way to do it. After you explain it, it's like, that makes sense. Of course that's how we should do it. And I really like the using of prior art maybe that isn't well-known, but seeing that diamond in the rough, I'm like, this is something that everyone can benefit from. That takes a lot of time.

Yeah, I mean, a lot of my projects now have been trying to focus on making them distro-agnostic. So while ClearBitManager obviously started out with ClearLinux to solve a very specific problem, the problem there is we didn't have packages. We had bundles. So you don't have this benefit anymore of, you know, having these meta packages and these new dependencies being introduced.

You don't have that. You have a static file system contents that's atomically written to this. This is what you have to work with. So the idea was to turn the update mechanism into the delivery mechanism. So you'll have like these user lib kernel files that turn up. The only thing that really changes there is the symlink. So you'd have like default LTS or default KVM. Those point to the newest kernel.

So there's none of this trying to figure out by the release number or the version which is the newest thing because i did that in the past it doesn't work very well so it looks at the symlink what whatever's the newest one is the default one which should be your basically your head or your tip kernel and it took a while to get that right but once it started getting fleshed out it's like this clearly could be used in other places damn so i set about making that agnostic and i've been trying to do that with a but once it started getting fleshed out, it's like, this clearly could be used in other places. Damn. So I set about making that agnostic, and I've been trying to do that with a lot of my projects. Yeah, you have. I agree.

I don't agree that everybody always takes it up. Sometimes somebody else, somebody else just creates the same idea, and But it's a start to try to have less of that. You think Arch would actually, would be open to something like this? Because

I mean Arch is the perfect candidate. I agree. They've got a very similar process to the way that Solus does it. You know like you have this kernel package and then you have this danger of you removing a live kernel

and live modules from the system. Yeah. And that's just incredibly dangerous. And it's got to lead it's got to be the number module breakage has got to be got to be the number one rolling breakage. It's got to be the number one rolling breakage cause, right? My USB no longer works because these modules have buggered off.

I can no longer use my speakers. You have all these problems. But something like this would be very easy for them, and it's a very fixed scheme. So you'd have a user lib kernel, and then you'd have like a kernel dash com.solus project blah blah blah each one has a a namespace i configure time so you customize it for your distro the the branding stuff that's put into the loader files that's just taken from the os release file so it's all standardized stuff one of my rules when i'm creating anything and you'll see this even in solos or any other projects i, not a lot of what I do is original or new.

It's just doing something in a sensible way. I'm not necessarily inventing anything new because, again, this is using existing technology. It's just trying to make it sane basically. You know, I feel like – well, I guess I'll ask you this question. Do you think the success of the ClearBoot Manager might be dependent on the success of ClearLinux? Do you think it stands on its own? Because to me it feels like it stands on its own.

Well, I mean, it's not one, I mean, I'm not going to talk about the success or whatever of ClearLinux because obviously I'm not in any capacity to do that. But ClearBoot Manager itself, while it started out as just a solution for ClearLinux, is something that of clear linux because obviously i'm not in incapacity to do that right but clear bit manager itself while it started out as just a solution for clear is something that quit very quickly became obvious it had to be agnostic and you know that's something that's maintained in my own namespace up on github quite deliberately and at some stage i probably will change the name as

well because it was at the time it's like okay what's a generic descriptive name all right clear bit manager yeah i actually i really like it i think it's actually i think it i think it's very good i we're going to get it adopted very very soon in solar so over the next week i mean i'm on vacation at the moment but i'm going to be finishing this up over the next week along with the the other Solus bits. You know how vacations work. You just go mad at projects. So I'll get that finished up, tested. Once that's in, we can just throw that straight into Solus. And then immediately everyone is then on...

I mean, it's been an issue for over a year in Solus. People are like, I update, and then my Windows entries are gone because what actually happens there is your NTFS and FAT modules have disappeared. because it's the first time it's trying to load them using os prober it tries to mount the file systems it can't do them because the modules are gone so you lose windows from grub and those are the sort of usability problems we've had or i've stuck in my headphones they're no longer working yet sorry you're gonna have to reboot you know the whole

thing sounds it sounds like it sounds like a really great solution. And I can't wait for Solus to ship it because that's it's an example of something that Solus is willing to do that. I just don't think Ubuntu is going to do right now or Fedora is going to do right now. Maybe I could tell. But I think once Solus proves it, I think I could see them doing it. Maybe I could tell – but I think once Solis proves it, I think I could see them doing it. But the reason why I mention that is because it takes me back to the conversation that Noah and I had months ago where he felt like he was kind of writing off Solis because it was sort of boutique. It was esoteric in a sense. It might be slightly boutique at the moment, but it is in that position where it's – you guys are the right size and have the right temperament to implement sweeping changes like this that people that are playing it a little bit safer are not willing to make.

And the issue is if you're always living in the little bit safer zone of Linux, you kind of miss out on some of the best stuff that's being – that's happening. And you kind of miss out on some of the new solutions to solve the problems. Like this could potentially one day legitimately lead to bulletproof kernel updates on Arch and Solus and other distributions. That's a big deal. That's a really big deal. And if you're hanging out on Ubuntu and it's fine, but you're going to miss it until it trickles down years down the road.

And that's totally fine and legitimate for a lot of people. But for some of us, if you're not riding that Solus wave or you're not on the Arch system or you're not – insert DistroX, right? Whatever. You're going to miss it. And I think you guys are in a really great position because you've got a good size user base. So it matters.

These things matter and you have to be careful. You have to contempl careful you have to contemplate very carefully about this but you're also not at the size where it's so so large that 200 people lose their job if everything falls apart and so i mean that's the thing we we don't have to play it safe and one of the things that we have that the other projects don't and it's not me blowing smoke up my own arse here right we're one of the very very few independent projects that is not commercial viability is not an aspect in any decision and that's something that we have that none of the other independent projects have because even though they might be independent and

they might be you know they might be 501c or whatever they still have interest to consider whereas we are fully independent we can go make changes. And there isn't a board to satisfy, there are no shareholders, they're to get again, frightened by the changes. So we can go in and make the changes that nobody else is willing or able to make. And then, you know, they can say that we've done it, well, it's all right. And then it might change their thinking as well. And that's one of the things that's it goes back to again like you know how do you measure success if we get people to rethink how they've been doing these things you know we're the disruptor at the moment when it comes to Linux we've chosen desktop Linux and we're going to

disrupt and make people reconsider the way things have been going on for the last 10, 20 years, whatever. So if we can only just do that, then we've already succeeded. I like that, how you phrase that, especially given how much we talk about Fedora as playing kind of a similar role of pushing the edge. But from a very, you know, Red Hat comes from, you know, it's a very traditional enterprise perspective. So I like that Solace can come from a very different place and innovate in the same space. That's a super good point.

Noah, were you about to jump in there? Yeah, I was just going to say, so I think saying I wrote Solus off might be a slight mischaracterization. So I think that really what I was trying to get at was that it's great to have people that are idealists, like Ike. And I think it's great to have projects that have idealists.

I would have to stop you there because I'm not an idealist. I agree. I'm very pragmatic. It's the complete opposite. If you're pragmatic, though, why exclude Java? Java is a very needed thing. Nobody likes Java. I don't like Java.

But the reality is I need Java to do half of the things I do during the day. Well, the last time that we tried to use it, the last time we tried to use Solus, Java was not working properly, and SimpleHelp didn't work properly. Oh, SimpleHelp. I remember that. But just Electron happens, and it's not even Java. No, it's not Electron. Here's the thing.

We can go on. I mean, I can pick it apart piece by piece, right? I could go on and say there's, you know, like we could go back to the Minicom thing. I need to be able to use serial communication. But my overall point is that you're making – you make decisions with the project that say we're not going to do X because I have a belief system of Y, and that's a good thing. And I think you should be commended for it. And I think you should be commended for it. I don't think you can necessarily say that we're on the same level playing field as somebody like Ubuntu who says, here are the requirements of XYZ users.

Here's what they want to be able to do. Let's accommodate that and make what compromises we need to make to make that distribution useful to those people. Sure, yeah. Yeah. All right, but what am I doing then if I'm not making compromises for users and addressing real-world needs to look at and reevaluate problems that have existed for a very long time in Linux? And I'm already looking at how other distributions do it. Would you characterize it? I'm not saying they're doing it wrong. I'm saying that they're doing it wrong. How do I add software to your distribution without compiling it from source?

Install it from the repos. If we don't have it, request it. Flatpak will be the future method. Okay, but what if... How do I do it right now? If it's not in the repos and I don't want to compile it from source, but I just want to install the software,

how do I get it there? Request it. You have a bug tracker. And in fact, we redid the requirements after that last, how should we say, mild fiasco. We've redid the requirements after that last how should we say mild fiasco we've redid the requirements because i basically said the same as everyone else the requirements were complete bullshit we know they were bullshit right but we hadn't had a discussion about it until that point

they've been but you still but you still have a gatekeeper you still have ultimately you still have a person that decides this can be right but that's no different to debian or ubuntu you still have a person that decides this can be... Right, but that's no different to Debian or Ubuntu. You still have to request a package that's not there. That's not entirely true. I can spin up my own PPA. Right, but somebody else has had to package it. How is it different from Solus when somebody has to package it? Could I not package it myself on Ubuntu?

Yes, but you're now going around in a circle from what you've just said. You've said if it's not on Solus, and I would then have to package it. That's a bad thing on Solus. But you're then allowed to package it on Ubuntu? What I'm saying is that if I was a software manufacturer, I could make my own package. I could make my own PPA. You could just make a dev file. Sure. I could make a dev file, put it up for download.

If you want PPA support on Solus, sure make it a dev file put it up look at this way if you want ppa support and solos ask me and i'll set it up i'm not an ogre off in the distance i'm willing to do things if there is demand for it i will happily do these things there has to be an actual demand but i guess my my my my i guess my overall point is the fact that we're having this discussion now rather like you didn't you didn't see that people would want to add software. And if you didn't have your own method, you wouldn't have included PPA. That's something that after we've already had the discussion of it being a batik or not batik distribution, now we're having the discussion of how we're going to maybe include PPAs. So if we go back to the core problem here is about software availability

right the thing is if you want something like ppas you also need people willing to maintain those ppas you know you can't just say i'm gonna put a ppa system up because i'm wasting money i'm pissing money in the window to do that right it doesn't help anyone we've got to a point now where we have people coming through and i think so far we've had something like 15,000 patches with multiple revisions come through to Solus from the community. We're at a point now where our tooling is in place, where we're switching our infrastructure over. There's actually an open bug for this to allow community maintainers

for packages within Solus. So that in itself isn't really a problem. There has to be people willing to package those things. Our packaging is very, very, very simple. I made our policies very, very simple for those to do that. Now, yes, the software isn't completely there at the moment, but then again, we're not exactly an old project. So there are going to always be things that you want and aren't there yet. If you want them, ask us, someone's going to put it in. It mightn't even be us now because people from the community are doing it. It takes time to build things up up but it takes even longer to build them up properly and that's what we've

been doing so would you say that you're so would you say your position has changed from the last time we had a conversation where you said that part of your part of your part of your methodology was to exclude software that you didn't think was appropriate or well maintained i've been more lax on it yes right because first of all we had a policy that's basically come through from the evolve os days and you have certain protective policies in place because at the time you know your capabilities you know your size and you know your your technology at the time our tooling very much sucked and we didn't even have y package or soul build back then so there are certain things we basically said absolutely no way in hell are going in because

it's going to cause us grief down the road we've now got to a point where y package soul build all of these tools exist and it's very very very trivial for anyone to just sit there and package something it's a yaml file you know we've a troop builder. You don't need to install a ton of dependencies on the system. It's very, very easy. So it's allowed me to be more lax on them. Did I take too long in doing the policy? Probably. But the conversation wasn't properly had.

We saw how it went. Someone went and bitched on Reddit. That's not how you do things. You know, people that know me know I'm very upfront. You know, I'll be very blunt about things, but going whinging behind me back on Reddit, it's not going to do anyone any People normally know I'm very upfront. I'll be very blunt about things, but go and whinge in behind me back on Reddit. It's not going to do anyone any favors. Raise it up with me, and we'll fix it. I would respectfully disagree with the characterization there as well.

It's one thing to me complaining on Reddit is simply going to Reddit before taking any other action. Right, but I hadn't seen it. This is the point. If it didn't go to me, how was I going to know? so where on the solace site does it tell me that i need to dev.solaceproject.com there's a tracker link in the bottom but my point is that he attempted to go through what you know based on based on his based on the time constraints that he had he went through to try to submit an issue to you guys he got formal communication back from the solace project maybe maybe you had or had not seen it,

but how is he supposed to know that, you know, what I'm saying is, like, if I submit something to Canonical, I don't wait for Mark to come back and say, you know, and say, did Mark personally review this? When I submit something to System76, I don't make sure that... No, no, no, no, no, no, no. You're making a very unfair comparison there, right?

We are not the size of Canonical. We do not have 700 employees fine fine compare it to altus i have five i have five the lead developer personally review every single issue that comes through our company i delegate out to other people and i would expect that a customer wouldn't expect me to review every single issue as well so i think that i think that i think that just because that you didn't personally see it to then and then after he took steps to submit it to you and then was rejected and after that process completed then went on reddit and said here was my experience i think to characterize that as complaining you know that might be a bit

unfair to him right but what i'm saying right long story short is could it complain then say look let's look at the policy nobody then said let's look at the policy and i've said we took too long in fixing the policy And I've said, we took too long in fixing the policy, but we did put a policy in place. It wasn't the right way to go around it. And that's all it comes down to. I suppose. And that's a matter of opinion. At the same time, I also don't think it's necessarily appropriate to expect your users to suggest policy revisions to the project. Because, you know, like to a large example, revisions to the project because, you know, like to a large example, as a user, if a software company or software project doesn't meet my needs or isn't willing to work with me, at some point I just say, okay, that's the direction they want to go. I'll go a different direction.

That's fine. I mean, look, everyone's entitled to their own opinion. But we are talking about someone who did provide a patch, right? And that does infer not just a user, right? So let's get that bumper out of the way. That is a contributor and a developer. Yes, they had a bad experience, but that was the policy at the time. And at the time, the policy said, no, this thing can't go in, right?

Which was fair because the policy then said, this thing is basically dead. It's abandoned, right? Then we did revise the policy afterwards. Yes, it wasn't exactly a great experience the way it turned out, and I don't think anyone wanted it to turn out that way. Me and that particular guy spoke and read it afterwards, and we said it didn't really

need to go this way. Right, and I followed that conversation. One minor correction, too. Chatroom is reminding me, it wasn't simple note that I had problems with. It was simple help, which heavily relies on java just one minor correction so yeah we've only got open jdk at the moment i guess you do like the oracle jdk so i just want to might be part of it i just want to step in here for a second and just say you know when i uh when i think about this uh i think what i what i what i do take away is uh almost every time Ihi comes on the show, he's evolved in a way that's logical and understandable on a position.

Like he just sprinkles on things like flat pack bombs and things like that in his conversation generally. But that's a huge – that is a huge change from where you started originally on this show. from where you started originally on this show. And as someone who also tries to allow myself to change my mind on things and just get into something and then decide to change it, because I feel like if I don't evolve my position, that I'm not taking in new information, I'm not taking in new input,

I'm not considering all of my possibilities if I just stay really fast on a particular course. So I guess I look at this and I think, well, screw all of this crap. Let's move forward and just get Flatpaks. Like if this – I'm sick. I am so sick of this conversation. Let's get Flatpaks up in here. And you know what?

I'll install the three programs I need that aren't in the repo at this point, maybe two at this point that aren't in the repo from Flatpaks. I'm good. I'm good. It comes down to a simple thing. It's, it's, the solo started off very, very, very, very tiny. It was basically just me originally making something for myself. It grew a lot quicker than I expected. You know, it's, it's had problems. It's had growing pains. And every so often, you know, we go's had problems it's had growing pains and every so often you know we go through this new expansion right now we're right on the edge of one and you always know when it's coming because you can feel the tension we have these kind of debates as well god do i know what you mean they do happen right and at the moment we're on the verge of expanding again it's like an

exponential growth we do it every few months right and the the problem we have right now and it's like an exponential growth. We do it every few months, right? And the problem we have right now, and it's basically a meme. Solus has no software, right? And yes, it's valid. But the thing is, we're willing to look at ourselves and say, right, this thing isn't being done right. We could do it better. So in terms of maintenance load

and core software availability, we're basically moving all of our infrastructure into Fabricator, which is the dev.solusproject.com. What that'll allow us to do is allow people within the community to own a Git repo within Solus, which corresponds directly to a package. They'll be able to publish it directly to the unstable repo. Obviously, we'll have checks and stuff, don't worry about that. But for the core repo, we're going to go from just a few gatekeepers. And that kind of is the problem here, isn't it? It's gatekeepers, which is what Noah said, right? We'll go from just having a few, select few, to basically this tick list of what is okay to go in.

And we've even got a two distro rule. If it's in two mainstream distros, you know, it goes in. Interesting. Interesting for sure. Because it was so anal, but it protected us during growth so then we'll have this checklist it's there's a few things that it basically says you know like if it's introducing huge stack complexity whoever said getting there is half the fun never whoa whoa kyra sorry i'm over eager kyra was very excited about you. Huge. That was really scary. That was one of the scariest things that ever happened on this show, actually.

If it introduces really large complexity, like something like Plasma, then someone would have to maintain it because we can't do it ourselves. That's very logical. Outside of the core applications, we're going to back Flatpak. We're going to help get them hosted. And we've done the the gl validation for the the open gl driver support nvidia and stuff so there will be choice it's not all completely there now but if you look at it the the last year or so we've been building that engine

that core part of solos and defining what solos actually is right now it's the time when we say let's flesh it out let's legitimize ourselves as a distro. And that's where we're at now. So we've been a project. We've focused on the OS aspect for a very long time. Now it's about legitimizing ourselves as a distro, as a larger player within the scene. And yeah, we're going to want the community to help us do that. You know, it's like Kairos is.

Getting there is half the fun. Exactly. It's half the fun. Just like people are watching L go through a a transition uh you know we watch solace in some sense go through a transition and uh i i don't know i so i find that all to be i find i find everything from from flatbacks to clear boot manager to the latest mate all to be extremely interesting development so thank you for coming on i would say one last thing before i go now a lot of things i say are a lot of things i say about solace i am not saying in a blanket statement way solace is for some people it's not for everyone it's the same as for a budgie it's

for some people it's not for everyone when i've complained about things like valor before it's it's good for certain jobs it's not good for everything right use the right tool for the job we need to stop being so religious and trying to blanket apply these things to every given situation solace is not trying to dominate every market it is trying to go for the home and give users a good experience that does not mean we're going on satellites and we're going on servers or raspberry pies we've a given focus if it doesn't work for you that's great we work on projects that go on over distributions to work with us on those things i like that boy that's a great philosophy i got

one question for you uh bungee 11 when am i getting it bungee 11 when's this shit i know when is it should be right right yeah we're doing 10.3 first because we have some issues with 3.22. That's old stuff. I don't want it. Yeah, it's old. It's crusty. It's well refined. It works good.

It's stable. Well, apart from all the time. Who wants those? Don't even get me started on the old. Okay. So, Budgie 11, we're looking to have that basically in next quarter. We're right on the edge of going into next quarter now. And we will be wanting testing on that.

And I'm going to try and see if we can do some kind of unstable testing images. So you can just download them and give them a go. Don't know how that's going to be working yet. But, yeah, I mean, that would be kind of cool if you can just get, like, unstable builds of Budgie and try it out. Yeah, I think that we can maybe submit a few if you can just get, like, unstable builds of Budgie and try it out. Yeah, I think that would be – and maybe submit a few bios. It's going to be a few months yet. Very good. Very good.

Yeah. All right. Kind of stay tuned. Well, thank you very much, sir. Thank you for taking the time. And like Kairos is, getting there is half the journey and half the fun. I know that. I'm going to go to Texas.

You're about to find out. You know what? Dylan, my son, my drive down to Texas lands right on a spring break. So he's going to go with us. He'll be there the whole trip. It's going to be a ton of fun. And you know what I'm going to do? I'm going to use my Ting MiFi.

I'll have a CDMA and GSM MiFi. Go to linux.ting.com to sign up. You get a $25 discount if you sign up. Or if you're bringing advice, they give you $25 in service credit. Now, your average Ting bill, $23. Can you believe that? I heard an ad for a network the other day. It was just $23 – or no, actually it was $22.

It was just $22 for the entire line. That's before your usage. That's before – $22 just to add a line. A line. $22 just to add a line. A line, $22. Ting is $6. GSM and CDMA, $6. $22 just for the line. So it's $6 and you pay for what you use.

Minutes, messages, megabytes. You add them up, whatever you use, that's what you pay. It's really easy to keep track of too because they got apps for your mobile device and they have a totally killer website. They also have really good customer service if you need to talk to a human being and get some actual support from a geek. Ting's got that. They have CDMA and GSM so you can either bring a device or pick whatever works best for you.

I'm mentioning that because that's huge for me right now because most of my trip I'll be on GSM. But there will definitely, absolutely, without question be times where I need a CDMA network. And that's what's so great about Ting. I got the tools to keep control. I can turn them off when I'm done. That's something else I think that is maybe overlooked because there's no contract. There's no determination fee. You can turn a line on for two months and then turn it off.

I got this little tracker that I've been thinking about putting in the rig again so people can track the trip. Well, I use that once a year. Am I seriously going to go get a contract or – I'm sorry. It's not an agreement. Am I going to – for something I use once a year? That's ridiculous. I'll put it on the CDMA network. I'll pay it for one month and then I'm done.

I can use these devices like the tools they are. Linux.ting.com. Go there, check them out. And you know what? Since we had a little random Kyra, let's go get an app pick from Kyra. Because otherwise, that's just like we got hanging Kairas. So we gotta get this in here. Whoever said getting there is half the fun never flew economy.

Oh! I'm Kyra and this is the Ting App of the Week. Take the RV, Kyra! Tripcase? Tripcase aims to take the trouble out of travel. If they don't already have a slogan, they're welcome to use that one. The first time you try Tripcase, it might blow your mind. When you get a travel email confirmation, just forward it to trips at tripcase.com. That's it.

All the relevant details of your trip are now on your phone. If there's a change in your itinerary, if your flight is delayed, or the gate changes, you'll be among the first to know. Tripcase can send you a push notification of changes if you give it the okay. It's about more than air travel, though. Tripcase can help you plan your trip on the ground, too. See details on your hotel check-in and out as part of your itinerary. See what the weather's like at your destination with a 10-day forecast.

Is that perfect or what? TripCase has detailed maps of many airports, which is super handy, especially if you have time to kill before you fly up. All right. Decide what you want to do while you're visiting by pinning attractions, appointments, and stuff you want to see on a map.

You can even get driving directions or book a ride with Uber with a couple of taps. Damn, dude. TripCase is free in the Apple App Store and in the Google Play Store. All right. Links are directly below. Thanks, Kyra. Thanks for watching. Subscribe to Ting True cases free in the Apple App Store and in the Google Play Store. All right. Links are directly below. Thanks, Kyra.

Thanks for watching. Subscribe to Ting on YouTube and get the latest App of the Week episodes, quick unboxings, reviews, and much more. Ow! Until next time. Linux.ting.com. It's mobile that makes sense.

Thank you to Ting for sponsoring the Unplugged program. Now let's roll into a few more updates, just really quick. First one I like a lot here. This is a big one for the beard. Serious Sam Fusion 2017 is rolling into public beta. He says this is a great game. You ever played any Serious Sam? Oh yeah. I have not. Oh, actually that's not true.

I've played one of the Serious Sam. Also, there's going to be VR support and the other thing that's really big is Vulkan support. In fact, Core Team is bringing Vulkan to their Linux ports. This is another big publisher that's going to be using Vulkan.

Yeah, man. This is a huge deal, actually. So it's in beta right now, 64-bit Vulkan, multi-threaded rendering, totally new save system, all of that. It's just buzzwords, man. All the buzzwords. Buzzwords I love. Also, while we're in the sort of media consumption category,

this is no longer a Valve update, so we'll close the Valve section. Goodbye. Mozilla has proposed Obsidian. Do you know what Obsidian is? I don't. Well, besides being obviously the original user. Hold on. Hold on a second. Hold on, Wes.

You got this? We can't... I gotta get these out of my pajamas, but I feel like we can't go any further. I feel like I would be negligent in my duties as a host of the show. Define obsidian. Obsidian is a hard, dark, glass-like volcanic rock formed by the rapid solidification of lava

without crystallization obviously i think the irc is right though we can't forget about the obsidian order uh yes right right the obsidian order is uh one of my favorite uh groups of uh spies that are directly modeled after our own c, which is pretty frightening. So Obsidian has been published as a possible proposal from Mozilla as the next WebGL technology. Yeah, as something – you probably stay hip to this, but the Kronos Group has been getting ready for WebGL next. And they've been looking for new proposals on the next version of WebGL. And guess who one of the first to submit a proposal was? Mozilla Foundation. They say this is a draft proposal

for the new GPU API for the web called Obsidian. It's a low-level API that provides maximum feature set for the GPU to web applications. The API is designed for WebAssembly, modern GPUs, and multi-threaded environments. Now, WebAssembly just recently shipped in Firefox and multi-threaded environments. Now, WebAssembly just recently shipped in Firefox as a default feature.

It's kind of a big deal. Also shipping in Chrome's latest, I think, version 53. And it's in WebKit's development builds, but not in any, like, Safari or anything. Now, Mozilla already has a prototype that's been built into Servo, which provides a metal-like API in WebIDL for JavaScript backed by Vulkan. Yeah. So this is a big deal. This Obsidian thing could work out to provide –

That's one of the reasons I'm glad Mozilla is still around. Yeah. Yeah, because otherwise who knows who would be pushing this agenda. Yeah, exactly. Now, I also wonder why we're in the media section. And I know we have a big conversation still to get into, but I wanted to talk about this really quick, because I think this is going to be

one of those things that silently passes in the night, but has long-term ramifications. At midnight on March 20th, 2017, Dolby's last relevant patent on Dolby Digital expired. So we're talking AC3 here.

So if you have AC3 audio in any television show you've ever captured, any television show or movie you've ever downloaded. In fact, some cameras record in AC3 audio. All digital television is – it's mandatory actually for digital television. A lot of internet streams actually come in AC3 that are high fidelity. Netflix can use AC3 to give you a surround sound. Yeah. And if you take out all references to Dolby and any Dolby properties, you can now ship AC3 technically.

This is huge. It's out of patent now. Yeah. And you can find out more at AC3, the number three, freedomday.org. And if you – first of all, if you've ever played for the Fluendo media pack like I have a couple of times you were paying for AC3 that's going to go away now that's just you can

just be included in the distro now I would assume and it also means that most live TV content can now be legally decoded most recorded TV content from like a PVR legally decoded videos from DVD or Blu-ray the audio can be legally decoded now. Some camcorders and mostly Netflix and maybe Hulu that uses AC3 in their streams.

I'm not sure. And maybe Amazon. Anybody that offers surround sound. But you see it a lot in established media. Oh, yeah, man. Yeah, it's all over the place. Pretty much all of the videos I play back on my media box probably have AC3

audio. Almost all of them. So that's a big deal. And again, you can find a link in the show notes if you want to check it out. It's funny because it's one of those things where one day, you don't even notice it, right? It's just eventually distros will start, well, it's been six months,

they haven't sued anybody, let's just include it. And then it just, all of a sudden, you're just legally playing files you didn't even realize. It's like it used to be back back in my day, we couldn't even play MP3 files, or

MP4 files. You couldn't even play an H.264 file. And now, the kids today, they're just randomly playing MP3 files. You couldn't even play an H.264 file. Yeah, right? And now, the kids today, they're just randomly playing MP3 files or randomly playing H.264 files, and now, their AC3 audio is just going to work like nothing. They didn't have to buy a codec pack. They didn't have to extract

it and put it into a.folder. Kids today. Yeah, there was no sketchy G-Streamer, bad, ugly, ah! Right. Yes. Oh my god. Oh, my God. Oh, man. All right. So there's a topic that I've been wanting to discuss, and I'm really glad that Noah's here because I think he's going to be somebody that's going to chime in on this.

But before we go, I got to do this because even just recently, just recently, spun up a DigitalOcean droplet, tried out an open source project, did a little hmm and ha, and then just destroyed the droplet, and probably in two hours I was done. At DigitalOcean.com, go there and use our promo code DEOUNPLUG to support the show. This is a great way to get a Linux rig with a crazy great 40 gigabyte e-connection, all SSD storage, KVM for the virtualizer, with data centers in New York, San Fran, Singapore, Amsterdam, London, Toronto, Germany, and Dimension B. They got them everywhere. You have a great control panel that's really easy to use, super intuitive. If you're a total noob that just barely kind of understands what a server is, maybe you've never even heard the term VPS before or Droplet, you're going to be able to use this. If you've been setting up virtual servers for a decade, you're going to be able to use this.

And best, this is like, do you use the API? Do you? Oh, yeah, totally. Oh, man. That's me throwing a piece of paper around because I'm so excited about the API. For me, I use it every single day, multiple times a day. What kind of things do you use it for? Well, my favorite part about it, one, you can just use it like, I mean, I've used it

in my own scripts, but because it's so easy, a ton of community stuff like Vagrant, right? Yes, dude. Vagrant has a writer. Basically, any scripting component, they probably have a drop-in digital notion. You spin up a droplet, you run your code there. Done. A couple of open source projects I've checked out have a button on their website that uses the API, and you click the button, and then it just connects to your account and spins up a droplet with the whole stack.

Ubuntu LTS, Docker, a container. I mean, it's just like one button and you have the whole thing. And I feel like D.O. is up there, right? Like even for some of the more maybe quote-unquote enterprise providers, you'll get a D.O. plug-in before you're going to get one of those. Yeah, and part of that too is they have incredible pricing. So it starts at $5 a month, but you can pay hourly, too, which is nuts. And then on top, they've layered on these incredible services that are integrated right into the dashboard. They're part of the API. They're first class, like load balancers, block storage, high RAM systems.

Like it's every move they have made has been really smart. And it's, I think, why they've just rocketed to the top. DigitalOcean.com. Use our promo code DL1PLUG. That lets them know you heard about it here, which means they keep us going. But also, it gives you a $10 credit. Try out the $5 rig, or my favorite, the $0.03 an hour rig. With a $10 credit, it really makes it.

There's several projects you could try with that. DigitalOcean.com. Give it a go. Build your back-end infrastructure, your future website, your personal blog, or just toy around and learn more at DigitalOcean.com. D-E-O-1 plug is the promo code you use after you sign up. Let's do a little elephant in the room discussion. There is a real serious reality. It's not something that we're totally shy of, but there's a real serious reality that for desktop Linux to be relevant and be successful long term, we're just going to have to have a whole crap ton of web apps.

Universal apps, maybe some are WebAssembly, some are Electron, just a desktop full of them. Simple Note, Slack, Skype, everyday business tools, web applications now. And it's just going to – everybody has their pet favorites. It's just going to increase, right? And I thought this was kind of an interesting blog post about setting a,

just a, let's just embrace this and set a.webapp file standard. So it's a file type.webapp and it's a free desktop specification. And if you use something like Natifier, which we're going to talk about a little bit more here in a second, this would be a way to just make a real simple, here's

the.web app, this is where you create the menu structure for it, this is the icon, all this stuff. It's like a known standard. And in part, when I read this, I was like, oh, geez, this is not the direction I want things to go. But then I thought about it, and it's like,

you have tools like Natifier. You have people that are distributing their entire applications in Electron like N1 is one that I still use today. I mean how many people use Atom or Visual Studio Code every day? Yeah, absolutely. And in some ways, this is critical for future expansion of the desktop I think for Linux because it's platform neutralizing. But we all are like intimately aware of what the problems are.

Many of them are closed source, completely closed source. Many of them require some sort of back-end infrastructure, which is also closed source and in some cases has pervasive user tracking that violates your privacy. And all of them so far seem to have a performance overhead when compared to native applications that is less than ideal. And so these are like obvious problems with web applications.

But then like at the same time, the advocate of me says, yeah, but screw all that stuff. I want Linux to be the successful prominent desktop platform one day. And you know, there's some advantages. As time goes on, I start to suspect that once again, we're never going to agree on a universal packaging format.

There's going to be app images and flat packs and snaps and we're never going to have a universal packaging format. It's never going to really be something that all distros support. It's never going to happen. But you know

what is universal? The web. So, why not bypass all this bullshit, all this crap that the Linux community can't get their shit together around, and just deliver a web app? Plus, large industrial vendors like Adobe and AutoCAD and others,

especially with things like Mozilla's initiative around WebGL Next. There could be like legitimate performance possibilities in an Electron app, in a web app. And they could just bypass.exe. They can bypass DMGs. They can bypass Flatpaks. Here's the web app. You can download it for these different systems.

Here's the web app. Flat packs. Here's the web app. You can download it for these different systems. Here's the web app. And then plus the other benefit would be, of course, that it's available to a shit ton more developers and people can target many platforms. So it's much easier to carve out some revenue because you can get a couple of cheapo Mac users to pay. You can get a few tens of thousands of Windows users and a few Linux enthusiasts to pay.

And so you're not tied to one platform to make your money. So when Microsoft finally kills the desktop and makes it just a Linux-based OS and Apple makes iOS for the desktop, you're not completely screwed. So you've diversified your income sources. So it's obvious for developers too because you bypass distribution standards. You bypass app stores. You bypass packaging issues. You bypass dependent on a platform. bypass packaging issues, you bypass dependent on a platform. It means like if I'm Adobe and I'm going to commit the next 15 years of development to something,

it almost seems like a foregone conclusion, which means as Linux users, we have to accept that our future is going to be tied to back-end services that require monthly payments, invade our privacy, and web applications that don't run as fast as they should, even though we now have these badass kit car style systems with an entirely open source platform. And we're running these closed source applications connected to closed source backends that run like shit. Is this a reality that we have to face? And is there some sort of middle in here that I'm not seeing? What do you think, Wes?

And after that, I'd love to toss it around the chat room. Joe's in here. I'd love to toss it around the chat room. Joe's in here. I'd love to toss it to Joe, see what he thinks. Colonel Linux is in here. I'd like to get his thoughts, but I'll let you start. I think there's a lot of things there tied together, especially the proprietary aspect. We're going to have proprietary software. We do already either way.

I think there's a legitimate case that's frequently made on this network that we do want that, that we want that as part of our success. Whether or not an individual Linux user wants that or not may be different, right? We all have opinions there. I think there is a good case that, you know, that we are tied to these backend services. But, again, when you get Google Drive integration in files on GNOME, that's the same thing, right? Like, you're still tied to proprietary service. What the frontend there doesn't really matter.

Good point. Where I see it from the development side, I agree. Like, there's some things that it makes sense for and some not. I'm running hyper.is, so it's a terminal written in Electron. I don't know if that makes sense. Really? It's pretty snappy, honestly. By a batch terminal?

Yes. I mean, a generic, you know, you can run whatever shell you want in it. Really? Yeah. Here's a shell. It feels very snappy. I only just installed it. It looks good, too.

It's in an app image. It does let you then have the full HTML if you want. My mind is spinning right now. And that's on the edge for me. I don't know if that makes sense. And I wouldn't necessarily do a tiny little app. You don't need a full Chrome thing just to draw a little bit. But on the flip side, if you've spent your career developing interactive web apps, you're using D3, you're using other things, are you really

going to go spend your time learning GTK? No. Maybe Qt. I think Qt has the strongest case, right, in terms of cross-platform availability. But where Electron shines is because it's JavaScript and it's web, and that is Reach. And if you're a developer with limited time, you're not excited about Linux or Windows or Mac. You're excited about the product you're making. Right, but okay, so let me ask Joe this. So Joe, what about the users though? Because like one of the reasons I switched to Linux is because I was sick of bad performance on other systems.

One of the reasons I switched to Linux is because I want a transparency of code. One of the reasons I switched to Linux is because I wanted long-term visibility of where a project was going. And what's the point of having a GPL kernel and GPL user land tools in a GPL desktop environment if I'm running a closed source application?

Is this a compromise too far, Joe? Well, I think it's not a compromise too far for most people because most pragmatic people will use a bit of proprietary software on the desktop regardless. But I think that you're kind of looking at this from the wrong angle here. You see, you're talking about these kind of consumer applications like Skype. Although I suppose Skype is a bit of a business application. But you see, the thing is the desktop is more or less dead or at least dying.

And everyone is moving to mobile, right? Except for the professionals using big software like Pro Tools. People that want to get their work done. Yeah, and that kind of stuff, it's very, very difficult to make that into web apps, isn't it? Well, I think today it is, but I don't know about long term. And so you think that maybe long term that is the way to get those big video editors and stuff over to Linux?

God, I hope so, in some ways. Because otherwise, it's never going to happen. I really think the only realistic shot that Linux has is that Adobe releases Premiere for Linux at this point. And the only way they're going to do that is if there's something that is substantially better to move to that is like the next 10 years plus for their platform, I would think. Otherwise, why not just keep releasing native builds for Windows and Mac? The other benefit maybe could be that it does make it, right, if you can get better web acceleration use of those hardware components, other things that you need that differentiate a web app from a real app, then you've solved it for everyone forever. Well, if you've got something like Obsidian and you start getting like a standard for audio and you start getting a standard for local database cache and stuff like that, you get to a point where you essentially have the workings of anything you could have made out of a Java app or anything you could have made out of something else.

It's in a runtime. I feel like you get actually pretty damn close. What you're talking about is the web browser becoming the operating system. Yeah, in a sense. is the web browser becoming the operating system. Yeah, in a sense. Or an Electron app or a Natify or something like that where it's using those technologies, but it pretends to be a desktop application.

And I don't like it, really. And in fact, see, Current Linux, I wanted to ask you because it seems like a huge component to all of these is then they always, it's not just that it's a web, or an Electron app, it's also that it's cloud dependent. And you have the whole lock-in now to a cloud service, which has all kinds of long-term ramifications, which could be one of them being shut down. And so this could be another issue where all of a sudden, one of the benefits

of desktop Linux was with open source native applications, a developer might decide to quit one day. But if your business depended on it, hey, you fork the shit out of that thing, run an internal build and hire somebody to keep it going. And that was an actual possibility. But right now, if Nalaeus decides to pull the plug on N1, I'm totally screwed. My desktop application becomes completely non-functional. And that's got to bother you, Colonel Linux. And additionally to that, when you start looking at independent content creators, now we're taking our content, things that we've spent our money, time, and effort, and we're putting that up into the cloud. And a lot of us would say, well, in the case of video editing and stuff like that, especially for Adobe, they wouldn't do anything terrible with it. But a lot of times what you see is as they change to accommodate users that don't want to pay a lot of money.

So, for example, on Facebook, when it started out, you uploaded a bunch of media. That was all your media. It was your pictures, your videos, that kind of thing. Then later on, they started changing the user agreement so that they could use some of that stuff. And so anytime you have content that you've created on somebody else's computer, I think that's a dangerous thing. Anytime you're running your business based on someone else's computer, I think that's a dangerous thing. Anytime you're running your business based on someone else's computer, I think that's a dangerous thing. So, I don't have a real way to

spin this into a happy-go-lucky hurrah desktop Linux thing. I guess I would just like to point out that you can have an entirely open source Electron app. The whole framework is open source. If you want to improve it to not use as much memory or whatever, and you're capable of doing that, you totally can. You can also use as much memory or whatever and you're capable of doing that, you totally can.

You can also use Electron with offline apps. You don't need to be online either. Yeah. Now, producer Michael, I wanted to give you a chance also to sort of touch on your thoughts on using Natifier to take some of these applications and make it essentially feel like a local app because that's sort of, I mean, where we're going. Well, I mean I've been using Natifier for a little while, and I've built like probably 10 of my own apps with it. It can do a lot of – it's really interesting because its purpose is to kind of like auto-build an Electron app for you. You give it – tell it what – you have a node and NPM installed,

and you have Electron in your system, and then it just uses Electron to build everything for you based on the parameters you use in the command. And then it just builds it and it gives you a full folder that's portable, but it also integrates with your system so it uses your home folder to store the config files so you can remake a new version with a new Electron when it comes out.

You can have notifications built into your, pretty much whatever DE you want. There's a lot of great tools for it and the best part about it is that you can build your own apps without having to deal with

any kind of services. I have stuff that's on my server that is a cloud service that is running on my server that is, you know, it is a cloud service that is running on my server that is using Electron to use it. That's cool. All right. So that's kind of, that's sort of like a, that's a middle ground. That's good.

That's good. That's a middle ground I was looking for. We're less depressed than we thought we might be. Thank you, Producer Michael. You gave us the middle ground that we were hoping for. All right. Now, if you have a little middle ground and you want to share your thoughts, go to jupiterbroadcasting.com slash contact!

Of course, you can also go to linuxactionshow.reddit.com although bit of a shit show right now, so I probably wouldn't. Maybe just find us on Twitter or join us live. Yeah, that'd be good. jblive.tv for the live show.

I'm at chrislas. I'm at westpain. The network's at jupiter signal. Also, that'd be good. JBLive.tv for the live show. I'm at Chris LAS. I'm at Wes Payne. The network's at Jupiter Signal. Also, go follow Colonel Linux at Colonel Linux on the network. He's over there doing his thing and tweeting about all his shenanigans. And also, we appreciate the Mumble Room for joining us.

You can also join that. Just check out the chatroom at JBLive.tv. Do a bang mumble, get the info. We do a little mic check, and you can join us. We'll see you right back here next Tuesday. And the value of this is negative. JBtitles.com. Everybody go over there. We'll pick our title

and then we will get out of here for the TechSnap crew. Thank you for being here everybody. JBtitles.com Also, I just want to point out another web app. Etcher is awesome. Oh yeah. I forgot about Etcher. Damn, that is actually the counter example right there.

Etcher for the win. And I've actually been studying up on app images with electron-based stuff. And it's really cool. I like app images a lot more than I used to because after researching it, I kind of like app images more than Flatpaks now. The first time I used Etcher, that was the first time i used uh app image and i was just i was blown away like it was it was one of those moments where the thing downloads i double click on it and i click on install and i'm like why why is not every application on linux using this and why haven't we been doing it for 20 years all right well because i i did it on my

arch desktop then i did it on my fedora laptop then I did it on my Fedora laptop, then I did it on my Ubuntu laptop, and the same freaking file runs on every single distro. Sure, sure, sure, sure. We need to pick a title, JBTitles.com. Sure, we need to pick a title, JBTitles.com. Sure, we need to pick a title, JBTitles.com. Chris loves it. That's funny. Sandboxing now, too.

Oh, really? Pack attack, Solus my Solus. Don't forget the limitations with app images as well. Of course, but there's also limitations with flatpacks as well. Well, not the same limitations. With an app image, you'll run into library issues. Oh! Yeah.

So they might bundle some of the libraries, but you're still going to run into the sort of issues that you would have if you just downloaded the terrible and run it from that root. Whereas if you use something like Flatpak, you've got a fully contained rootFS essentially. You've got an entire troop environment. No, because

AppImage has changed how they do their stuff within like four or five months ago. They're now using SquashFS now. So there's no... Yeah, right. But that SquashFS, that's just a bind mount that's gone into the troop. It doesn't have a full troop, so the system libraries can still contaminate. That's one of the problems.

AppImage is very, very attractive because you can just double-click it and go for it, as Noah said. But you can still get those issues, so they do need to be watched out for. If you use a sandboxing, that's not going to happen. Like FireJail, for example. FireJail on type of AppImage, you're not going to have those problems. Or you could just use one that works out of the box without layers. Right, but when you have to install Flatpak into your system and then your system doesn't work in any way

and then it gives you these ridiculously obscure errors and the actual solution is to reboot your machine, but the system doesn't tell you that because that'd be too easy. You had the Quark errors as well then. Yeah, it's ridiculous. It's like, hey, you're missing this thing. Like, why am I missing it? Oh, because I didn't reboot.

Oh, yeah, thanks. You could have just told me I need to reboot. Yeah, Snap kind of has the usability upper hand there a little bit. Yeah. But there's a thing. i think the app image is only there's it's only flaw not really only flaw because the sandboxing is a problem but they have a solution for that the only main thing that they don't have a solution for is the updating method like yeah yeah if they had a self updater that would be nice so they have a built-in updating

daemon but the app imageage has to implement it. And it only does it when you launch it. So it says, okay, here's an update. So it does work, but the AppImage developer has to use it. And it is really cool. Sounds like the Mac. Well, it is more Mac than not. But a lot of people think that AppImage just bundles everything,

and they don't. AppImage is only bundle what's necessary to bundle. If the system, if AppImage can assume that a package is available, like a super massive global package, it's not going to be bundled.